Voice recognition as a technique for interacting with user devices has grown popular. These systems typically depend on recognizing words and phrases spoken by a user. To recognize these words, the devices process received voice commands according to the language of the user. However, many users may speak in multiple languages and/or speak a language with a particular accent that may not be recognized by the user device. Moreover, the device may not be able to provide results that are specific to a particular language or accent.